Last time, we looked at the Runs Created in its basic version. In order to measure how many runs a hitter has created for his team, the product of his on-base-percentage and his total bases is taken (OBP*TB), or, if you like this better, the product of his on-base-percentage and slugging percentage is multiplied with his at-bats (OBP*SLG*AB).

Runs Created however has some flaws. First of all, it does not take stolen bases into account. Therefore, it did not take long for Bill James to create a stolen bases version of Runs Created. Since being caught stealing erases a hit or walk, it should decrease a players OBP and indeed, in the new formula, the OBP part changes from (BB + H) / (AB + BB) to (BB + H - CS) / (AB + BB). So far, so good. Now, the successfully stolen bases are a little trickier. The first idea would be to add them to a player’s total bases. But unlike hits, stolen bases do not move other runners further along, e.g. while they help scoring runs, they do not help cashing runners in. So maybe weighting stolen bases with one half is a good idea. Then again, stolen bases get you into scoring position and they take double plays away and this should count for something. In the “official” stolen base version of Runs Created, Bill James decided to weight stolen bases with the factor 0.55, only slightly higher than one half. Unfortunately, I can not explain how he came to this particular number, but he explains this in his 1984 Abstract (Source: Dan Agonistes), which, again unfortunately, is difficult to obtain outside the US.

Anyway, in the formula, stolen bases are added to total bases after being weighted with 0.55, which leaves us with the final “stolen bases” version of



(Now if this seems complicated to you, wait until we come to the next part.)

As you can imagine, including stolen bases was only the first step. After all, there are plenty other stats available like sacrifice flies, sacrifice bunts, grounded into double plays, intentional base-on-balls and hit-by-pitches. Even strike outs have a (negative) effect (we will later see why). However, as we have seen with stolen bases, it is not always easy to determine how to weight those stats and as a result, different versions of highly elaborate runs created formulas exist out there. Let me give you the technical version of Runs Created from 1988, which - although not the newest - is very useful to show a few interesting points:



Ok, this looks quite complicated, so we break it down into smaller parts. The former OBP-part now looks like this:

This is a bit larger than before, but still pretty intuitive. A hit-by-pitch pretty much works as a walk, while grounding into a double play removes a runner. Under the fractions stroke, we simply have plate appearances now after adding sacrifice bunts and flies.

Things get a little more interesting on the right side, the “slugging side”, if you will. In addition to total bases and stolen bases (weighed slightly different than before), we now have three other parts.

1) (BB+HP-IBB) * 0.24
At first, it is a bit surprising that walks and hit-by-pitches are found next to total bases now, but if there is already a men on first, these events do move runners along, apparently often enough to weight them with about a forth of a hit (0.24). Intentional walk however are subtracted, since they are usually only issued with first base empty and therefore, do not move runners over.

2) (SH + SF) * 0.5
Sacrifices do pretty much the same as stolen bases, e.g. moving runners up a base. In this formula, they are worth slightly less, but other versions of Runs Created weight them equally (Personally, at first sight, I thought you could make the case that the sacrifice bunt should weight more since they are often used to move two runners over instead of just one, but I have not really thought that idea through yet). If you wonder why there is no negative effect of the out made when sacrificing, remember that SF and SH are now found under the fraction line on the OBP-part of the formula, too.

3) SO * 0.03
Lastly, there are strike outs, which have a small negative effect because unlike ground outs and fly outs, they never move runners over. Again, personally, I would have expected the effect to be bigger since the “productive out” got so much press not that long ago when small ball was so hot.

Anyway, I hope separating the formula helped understanding the idea behind it (at least that’s what it did for me), although we are still pretty much left in the dark about where the different weights come from exactly.

Of course, this is not the last conclusion of wisdom. For example, the 2002 version of Runs Created has different weights for singles, doubles, triples and home runs and has a much more complicated structure (see the Wikipedia entry if you want to know the details). There is even a version that includes situational hitting. However, we are not going to get any deeper into this here now. Instead, I just like to get back to the fact mentioned in the last part: Runs Created is not very useful to compare the “inherent” talent of players. Runs Created “favors” hitters that play many games and hit at the top of the order, amassing many at-bats. To compare players, Runs Created needs to be expressed as a rate stat. Therefore, RC/27, i.e. Runs Created per 27 outs has been developed. It measures how many runs a hypothetical line-up of the same nine individuals (e.g. nine Chase Utleys) would score of average. To calculate RC/27, Runs Created is divided by the total number of outs (AB - H + CS + GDP + SH + SF) and then multiplied by 27.

For fun, lets check some stats from last season. You will see that there was only one player in the National League (or the Majors for that matter) who was even close to Albert Pujols when he took the field. (Stats are taken from fangraphs.com, who use a only slightly different RC version from the one described above)

These are the Senior Circuit Top 10 in Runs Created from 2008. As you can see, judging by RC, Albert Pujols was simply otherworldly and should run away with the MVP award if the world was fair. On the above list, only one man can give Albert a run for his money looking at RC/27. Make your guess now and then look at the following table:

Yes, a team of nine Chipper Joneses would score more than one whole run more than any fictional 2008 team with players not named Albert Pujols. Too bad he got injured after starting the season red-hot and chasing .400 for a while.

Ok, we are finished with Runs Created now, I hope you enjoyed diving a little deeper into the sabermetrican jungle and you are going to come back on our next trip behind the Mendoza line into the great unknown of modern baseball statistics.

Discussion (Add New Comment)

Write comment
Name:
Email (optional, not shown):	do not notifynotify
Title:
	Please input the anti-spam code that you can read in the image.

Powered by !JoomlaComment 3.26

3.26 Copyright (C) 2008 Compojoom.com / Copyright (C) 2007 Alain Georgette / Copyright (C) 2006 Frantisek Hliva. All rights reserved."

You can contact Bjoern Hartig via the writer's Profile or the AHP Staff via the contact form.

Hype up this post at BallHype!